Abstract
The dimensionality of the Rotter locus of control scale is discussed in terms of interpretability, comparison with other studies, the pattern of eigen values, and goodness of fit. It is concluded that the identification of more than two factors should be done with considerable reluctance.